Skip to content

feat(scheduler): export hami_resource_quota_limit gauge metric #2715

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
hami-robot merged 2 commits into from
Aug 20, 2026
Merged

feat(scheduler): export hami_resource_quota_limit gauge metric #2715

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9e15d2f
feat(scheduler): export hami_resource_quota_limit gauge metric
SK8-infi Aug 18, 2026
adf3113
fix(scheduler): omit hami_resource_quota_limit when quota limit is un…
SK8-infi Aug 18, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 11 additions & 1 deletion cmd/scheduler/metrics.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -281,13 +281,18 @@ func (cc ClusterManagerCollector) collectNodeMetrics(ch chan<- prometheus.Metric
}
}

// collectQuotaMetrics emits per-namespace resource quota usage.
// collectQuotaMetrics emits per-namespace resource quota usage and limits.
func (cc ClusterManagerCollector) collectQuotaMetrics(ch chan<- prometheus.Metric, legacy bool) {
quotaUsedDesc := prometheus.NewDesc(
"hami_resource_quota_used",
"resourcequota usage for a certain device",
[]string{"namespace", "quota_name", "limit"}, nil,
)
quotaLimitDesc := prometheus.NewDesc(
"hami_resource_quota_limit",
"resourcequota limit for a certain device",
Comment thread
archlitchi marked this conversation as resolved.
[]string{"namespace", "quota_name"}, nil,
)
var legacyQuotaUsed *prometheus.Desc
if legacy {
legacyQuotaUsed = prometheus.NewDesc(
Expand All @@ -301,6 +306,11 @@ func (cc ClusterManagerCollector) collectQuotaMetrics(ch chan<- prometheus.Metri
if err := sendMetric(ch, quotaUsedDesc, prometheus.GaugeValue, float64(q.Used), ns, quotaname, fmt.Sprint(q.Limit)); err != nil {
klog.V(4).Infof("Failed to send quotaUsedDesc metric: %v", err)
}
if q.LimitSet {
if err := sendMetric(ch, quotaLimitDesc, prometheus.GaugeValue, float64(q.Limit), ns, quotaname); err != nil {
klog.V(4).Infof("Failed to send quotaLimitDesc metric: %v", err)
}
}
if legacy {
sendLegacyMetric(ch, legacyQuotaUsed, prometheus.GaugeValue, float64(q.Used), ns, quotaname, fmt.Sprint(q.Limit))
}
Expand Down
46 changes: 46 additions & 0 deletions cmd/scheduler/metrics_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -438,6 +438,10 @@ func TestClusterManagerCollectorQuotaMetrics(t *testing.T) {
metricsProvider: provider,
}
want := `
# HELP hami_resource_quota_limit resourcequota limit for a certain device
# TYPE hami_resource_quota_limit gauge
hami_resource_quota_limit{namespace="team-a",quota_name="nvidia.com/gpucore"} 100
hami_resource_quota_limit{namespace="team-a",quota_name="nvidia.com/gpumem"} 8192
# HELP hami_resource_quota_used resourcequota usage for a certain device
# TYPE hami_resource_quota_used gauge
hami_resource_quota_used{limit="100",namespace="team-a",quota_name="nvidia.com/gpucore"} 50
Expand All @@ -447,6 +451,7 @@ hami_resource_quota_used{limit="8192",namespace="team-a",quota_name="nvidia.com/
collector,
strings.NewReader(want),
"hami_resource_quota_used",
"hami_resource_quota_limit",
); err != nil {
t.Fatalf("unexpected non-legacy collecting result:\n%s", err)
}
Expand All @@ -462,6 +467,10 @@ hami_resource_quota_used{limit="8192",namespace="team-a",quota_name="nvidia.com/
# TYPE QuotaUsed gauge
QuotaUsed{limit="100",quotaName="nvidia.com/gpucore",quotanamespace="team-a"} 50
QuotaUsed{limit="8192",quotaName="nvidia.com/gpumem",quotanamespace="team-a"} 4096
# HELP hami_resource_quota_limit resourcequota limit for a certain device
# TYPE hami_resource_quota_limit gauge
hami_resource_quota_limit{namespace="team-a",quota_name="nvidia.com/gpucore"} 100
hami_resource_quota_limit{namespace="team-a",quota_name="nvidia.com/gpumem"} 8192
# HELP hami_resource_quota_used resourcequota usage for a certain device
# TYPE hami_resource_quota_used gauge
hami_resource_quota_used{limit="100",namespace="team-a",quota_name="nvidia.com/gpucore"} 50
Expand All @@ -471,9 +480,46 @@ hami_resource_quota_used{limit="8192",namespace="team-a",quota_name="nvidia.com/
collector,
strings.NewReader(want),
"hami_resource_quota_used",
"hami_resource_quota_limit",
"QuotaUsed",
); err != nil {
t.Fatalf("unexpected legacy collecting result:\n%s", err)
}
})
}

func TestClusterManagerCollectorQuotaUnconfiguredLimit(t *testing.T) {
const (
ns = "team-b"
memName = "nvidia.com/gpumem"
)

unconfQm := device.NewQuotaManager()
unconfQm.Quotas[ns] = &device.DeviceQuota{
memName: &device.Quota{Used: 1024, Limit: 0, LimitSet: false},
}
t.Cleanup(func() { delete(unconfQm.Quotas, ns) })

unconfProvider := &fakeMetricsProvider{
nodeUsage: map[string]*schedulerpkg.NodeUsage{},
quotaManager: unconfQm,
podManager: device.NewPodManager(),
}
collector := ClusterManagerCollector{
ClusterManager: &ClusterManager{LegacyMetrics: false},
metricsProvider: unconfProvider,
}
want := `
# HELP hami_resource_quota_used resourcequota usage for a certain device
# TYPE hami_resource_quota_used gauge
hami_resource_quota_used{limit="0",namespace="team-b",quota_name="nvidia.com/gpumem"} 1024
`
if err := promtestutil.CollectAndCompare(
collector,
strings.NewReader(want),
"hami_resource_quota_used",
"hami_resource_quota_limit",
); err != nil {
t.Fatalf("unexpected unconfigured limit collecting result:\n%s", err)
}
}
Loading